As as Dev / Release Manager working on the next release of your product, how often are you asked the question “when will you be done”? And as a Product Manager or member of the executive team, how often do you get a satisfactory response?

Whether you’re asking or answering the question, you likely want the timeframe given to:

  • Be reasonably fast

  • Be provided with a high-level of confidence

  • Be relatively precise

For example, a timeframe of being ready in the next 6-24 months has so much variation that it is not reasonably fast, confident, or precise.

Though this question seems straight-forward, as a dev/release manager, I struggle to answer this question for the following reasons:

  • The scope was not clear enough

  • The scope may change between now and when we go to production

  • The solution involves technical risks / decisions that could significantly impact the timeline / effort of the release

  • There were a number of factors outside of my control

  • I do not know who will perform the work and their level or expertise

As a result of having so many unanswered questions, the typical response is: “it depends”.

Rather than providing the typical response, here are the ways I try to indicate “when we will be done” with reasonably fast, highly confident, and precise timelines:

  • Review high-level features and provide order of magnitude estimates.  If needed, I decompose the features into sufficient details in order to improve scope clarity and time estimates

  • Prioritize the scope to know which features are “must-haves” and which ones can be left out, if needed, to meet required timelines

  • Use Subject Matter Experts to create prototypes or spikes to reduce technical unknowns

  • Conduct technical readiness reviews to ensure timelines won’t get pushed out due to unexpected technical hurdles

  • Ensure the full team has taken part in providing the estimates that contribute to the timeline

    • If the team has worked together in the past, I have a baseline velocity as a starting point; alternatively, if I have a new team, I ensure I have a steady team composition and calibrate the velocity over 3-4 iterations

In conclusion, to provide a value-adding answer to the question “so, when are we done?”, consider the following:

  • What level of confidence and precision is required for your estimate?

  • How can you eliminate or control the variables that are going to impact your estimate the most?

  • What historical data can you use as basis for your answer: team velocity, estimates vs. actuals for similar projects, variation in the size of the backlog for a typical release etc.

How do you handle the “when are we done” question?

Posted by:Aequilibrium

With a track record for building and leading high performance Agile teams and delivering great software products and engaging user experiences, Adrian’s focus has always been on delivery excellence. Adrian is the Founder and CEO of Aequilibrium—a Vancouver-based digital product development and design agency dedicated to creating winning web, mobile and IoT solutions. Adrian also makes a mean espresso and can teach you a killer chess move or two... in English, French, Romanian or Italian.